RADT-30. MANAGEMENT OF PRESUMED PSEUDOPROGRESSION WITH EARLY BEVACIZUMAB IN THE THREE MONTHS AFTER HIGH DOSE RADIATION THERAPY FOR GLIOBLASTOMA

Abstract BACKGROUND Optimal management for symptomatic steroid-refractory pseudoprogression (rPsP) following IMRT for Glioblastoma remains poorly defined. This study examined radiological volumetric response and clinical outcomes following use of early Bevacizumab (earlyBEV), defined as commencing during or within 3 months of IMRT. METHODS Consecutive patients managed with Stupp60Gy IMRT between 2016-2023 for Glioblastoma were analysed for those patients receiving earlyBEV for rPsP. Volumetric analysis was performed with sequential MRI T1gd/T2 sequences using preBEV and postBEV scans. Primary endpoint was volumetric reduction of T1gd/T2 volumes at month+2/3 postBEV. Clinical response, overall survival and pattern of relapse, specifically distant failure was assessed. RESULTS Forty-one (14%) of 289 patients received earlyBEV for rPsP, with median follow-up of 9.9months (q1-3:8-25) for survivors. Compared to remaining 248 patient cohort, earlyBEV had worse initial ECOG0,1 (50%vs76%); less near-total resection (5%vs42%); and less MGMT methylation (35%vs46%). Additionally at diagnosis the initial tumours had larger T1gd and T2 volumes. Median time to commencement of earlyBEV was 1.4months postIMRT; including seven patients during IMRT. All patients responded to BEV, with reduction in symptoms, steroid use and radiological volume, though one died pre-assessment. Median T1gd and T2 volumes before earlyBEV were 37cm3 (q1-3:23-63cm3) and 116cm3 (q1-3:90-148cm3) respectively, which were 95% and 78% larger than at diagnosis. Median survival postBEV was 8.1months (95%CI:7.5-8.6). Median volumes postBEV reduced to 16cm3 (q1-3:8-23cm3) and 41cm3 (29-55cm3); which was a percentage reduction for T1gd and T2 of 57% and 65%. Midline shift reduced from 5mm(q1-3:3-8mm) to zero in two-thirds of patients. Patients requiring earlyBEV had worse survival compared to remaining cohort (p<0.001); with median OS 11.9months (95%CI:10.4-13.5) vs 19.9months (95%CI:18.4-21.5). A component of distant relapse was evident in 48% of earlyBEV vs 34% in remaining cohort. One serious adverse event occurred with fatal intracranial haemorrhage day10 after initial BEV dose. CONCLUSION This study data confirms efficacy of earlyBEV for managing steroid-refractory pseudoprogression, but also the worse prognosis of this high-risk cohort.